Enabling aircraft emergency landings using active visual site detection

The ability to automate forced landings in an emergency such as engine failure is an essential ability to improve the safety of Unmanned Aerial Vehicles operating in General Aviation airspace. By using active vision to detect safe landing zones below the aircraft, the reliability and safety of such systems is vastly improved by gathering up-to-the-minute information about the ground environment. This paper presents the Site Detection System, a methodology utilising a downward facing camera to analyse the ground environment in both 2D and 3D, detect safe landing sites and characterise them according to size, shape, slope and nearby obstacles. A methodology is presented showing the fusion of landing site detection from 2D imagery with a coarse Digital Elevation Map and dense 3D reconstructions using INS-aided Structure-from-Motion to improve accuracy. Results are presented from an experimental flight showing the precision/recall of landing sites in comparison to a hand-classified ground truth, and improved performance with the integration of 3D analysis from visual Structure-from-Motion.

Novelty-based visual obstacle detection in agriculture

This paper describes a novel obstacle detection system for autonomous robots in agricultural field environments that uses a novelty detector to inform stereo matching. Stereo vision alone erroneously detects obstacles in environments with ambiguous appearance and ground plane such as in broad-acre crop fields with harvested crop residue. The novelty detector estimates the probability density in image descriptor space and incorporates image-space positional understanding to identify potential regions for obstacle detection using dense stereo matching. The results demonstrate that the system is able to detect obstacles typical to a farm at day and night. This system was successfully used as the sole means of obstacle detection for an autonomous robot performing a long term two hour coverage task travelling 8.5 km.

From ImageNet to mining: Adapting visual object detection with minimal supervision

This paper presents visual detection and classification of light vehicles and personnel on a mine site.We capitalise on the rapid advances of ConvNet based object recognition but highlight that a naive black box approach results in a significant number of false positives. In particular, the lack of domain specific training data and the unique landscape in a mine site causes a high rate of errors. We exploit the abundance of background-only images to train a k-means classifier to complement the ConvNet. Furthermore, localisation of objects of interest and a reduction in computation is enabled through region proposals. Our system is tested on over 10km of real mine site data and we were able to detect both light vehicles and personnel. We show that the introduction of our background model can reduce the false positive rate by an order of magnitude.

MPC controlled multirotor with suspended slung load: System architecture and visual load detection

There is an increased interest in the use of Unmanned Aerial Vehicles for load transportation from environmental remote sensing to construction and parcel delivery. One of the main challenges is accurate control of the load position and trajectory. This paper presents an assessment of real flight trials for the control of an autonomous multi-rotor with a suspended slung load using only visual feedback to determine the load position. This method uses an onboard camera to take advantage of a common visual marker detection algorithm to robustly detect the load location. The load position is calculated using an onboard processor, and transmitted over a wireless network to a ground station integrating MATLAB/SIMULINK and Robotic Operating System (ROS) and a Model Predictive Controller (MPC) to control both the load and the UAV. To evaluate the system performance, the position of the load determined by the visual detection system in real flight is compared with data received by a motion tracking system. The multi-rotor position tracking performance is also analyzed by conducting flight trials using perfect load position data and data obtained only from the visual system. Results show very accurate estimation of the load position (~5% Offset) using only the visual system and demonstrate that the need for an external motion tracking system is not needed for this task.

Visual efficiency detection index : a new composite measure of visual search

Visual search is an important component of our interaction with our surroundings, allowing us to successfully identify external cues that impact our spatial navigation. Previous research has established fixation duration, fixation count, saccade velocity, and saccade amplitude as important indices of visual search. We examined the Visual Efficiency Detection Index (VEDI) comprising multiple aspects of visual search performance into a single measure of global visual performance. Forty participants, 10 adults ages 22-48, and children ages 6, 8, and 10, completed tests of working memory and visual search in response to stimuli relevant to pedestrian decision making. Results indicated VEDI statistically relates to established indices of visual search in relation to their interpretability for human performance. The VEDI was also sensitive to developmental differences in visual search performance, suggesting insight to its utility in the developmental psychological literature.

Laboratorial training of examiners for using a visual caries detection system in epidemiological surveys

Background: In epidemiological surveys, a good reliability among the examiners regarding the caries detection method is essential. However, training and calibrating those examiners is an arduous task because it involves several patients who are examined many times. To facilitate this step, we aimed to propose a laboratory methodology to simulate the examinations performed to detect caries lesions using the International Caries Detection and Assessment System (ICDAS) in epidemiological surveys. Methods: A benchmark examiner conducted all training sessions. A total of 67 exfoliated primary teeth, varying from sound to extensive cavitated, were set in seven arch models to simulate complete mouths in primary dentition. Sixteen examiners (graduate students) evaluated all surfaces of the teeth under illumination using buccal mirrors and ball-ended probe in two occasions, using only coronal primary caries scores of the ICDAS. As reference standard, two different examiners assessed the proximal surfaces by direct visual inspection, classifying them in sound, with non-cavitated or with cavitated lesions. After, teeth were sectioned in the bucco-lingual direction, and the examiners assessed the sections in stereomicroscope, classifying the occlusal and smooth surfaces according to lesion depth. Inter-examiner reproducibility was evaluated using weighted kappa. Sensitivities and specificities were calculated at two thresholds: all lesions and advanced lesions (cavitated lesions in proximal surfaces and lesions reaching the dentine in occlusal and smooth surfaces). Conclusion: The methodology purposed for training and calibration of several examiners designated for epidemiological surveys of dental caries in preschool children using the ICDAS is feasible, permitting the assessment of reliability and accuracy of the examiners previously to the survey´s development.

Effects of age and eccentricity on visual target detection

The aim of this study was to examine the effects of aging and target eccentricity on a visual search task comprising 30 images of everyday life projected into a hemisphere, realizing a ±90° visual field. The task performed binocularly allowed participants to freely move their eyes to scan images for an appearing target or distractor stimulus (presented at 10°; 30°, and 50° eccentricity). The distractor stimulus required no response, while the target stimulus required acknowledgment by pressing the response button. One hundred and seventeen healthy subjects (mean age = 49.63 years, SD = 17.40 years, age range 20–78 years) were studied. The results show that target detection performance decreases with age as well as with increasing eccentricity, especially for older subjects. Reaction time also increases with age and eccentricity, but in contrast to target detection, there is no interaction between age and eccentricity. Eye movement analysis showed that younger subjects exhibited a passive search strategy while older subjects exhibited an active search strategy probably as a compensation for their reduced peripheral detection performance.

Intelligence is related to specific processes in visual change detection: Fixed-links modeling of hit rate and reaction time

By means of fixed-links modeling, the present study identified different processes of visual short-term memory (VSTM) functioning and investigated how these processes are related to intelligence. We conducted an experiment where the participants were presented with a color change detection task. Task complexity was manipulated through varying the number of presented stimuli (set size). We collected hit rate and reaction time (RT) as indicators for the amount of information retained in VSTM and speed of VSTM scanning, respectively. Due to the impurity of these measures, however, the variability in hit rate and RT was assumed to consist not only of genuine variance due to individual differences in VSTM retention and VSTM scanning but also of other, non-experimental portions of variance. Therefore, we identified two qualitatively different types of components for both hit rate and RT: (1) non-experimental components representing processes that remained constant irrespective of set size and (2) experimental components reflecting processes that increased as a function of set size. For RT, intelligence was negatively associated with the non-experimental components, but was unrelated to the experimental components assumed to represent variability in VSTM scanning speed. This finding indicates that individual differences in basic processing speed, rather than in speed of VSTM scanning, differentiates between high- and low-intelligent individuals. For hit rate, the experimental component constituting individual differences in VSTM retention was positively related to intelligence. The non-experimental components of hit rate, representing variability in basal processes, however, were not associated with intelligence. By decomposing VSTM functioning into non-experimental and experimental components, significant associations with intelligence were revealed that otherwise might have been obscured.

Impact of Different Illumination Conditions on Visual Caries Detection with ICDAS

The aim of this study was to test the influence of different degrees of additional illumination on visual caries detection using the International Caries Detection and Assessment System (ICDAS). Two calibrated examiners assessed 139 occlusal surfaces of extracted permanent molars using a standard operation lamp with or without an additional headlamp providing three default brightness intensities. Histology served as the gold standard. Pooled data showed no differences in sensitivities. Specificities were not influenced by additional light. The area under the curve for the Marthaler classification D3 threshold was significantly lower when an additional strong headlamp was used (0.59 compared to 0.69-0.72 when reduced illumination intensities were used). One of the two examiners also had a significantly lower sensitivity for the D1 threshold when an additional headlamp was used. The use of additional white light led to a reduced detection of dentine lesions.

Impact of different magnification levels on visual caries detection with ICDAS

OBJECTIVES The aim of this in vitro study was to examine the effect of different levels of magnification on the accuracy and reliability of visual caries detection using ICDAS criteria. METHODS Occlusal surfaces of 100 extracted molars were assessed by 14 examiners (3rd and the 4th year dental students and dentists) using no magnification aids, a 2.5× Galilean loupe, a 4.5× Keplerian loupe, or a surgical microscope with 10× magnification. The assessments were repeated on a different day. Sensitivity, specificity, AUC and reliabilities were calculated according to the gold standard of histology. RESULTS We found that with increasing magnification, the number of surfaces rated as "sound" (ICDAS code 0) decreased, while the number of surfaces with a localized enamel breakdown (ICDAS code 3) increased. While the sensitivities increased, the values of the specificities decreased to an unacceptably low level irrespective of the clinical experience of the examiners. CONCLUSIONS ICDAS seems to be optimized for natural vision up to 2.0× magnification and not for high magnifications. The use of powerful magnification in visual caries detection involves the risk of unnecessary and premature invasive treatment. CLINICAL SIGNIFICANCE This paper discusses when it does and does not make sense to use magnification devices for visual caries detection using ICDAS criteria. Strong magnifications should be refrained from for this purpose.

HOG-Like gradient-based descriptor for visual vehicle detection

One of the main challenges for intelligent vehicles is the capability of detecting other vehicles in their environment, which constitute the main source of accidents. Specifically, many methods have been proposed in the literature for video-based vehicle detection. Most of them perform supervised classification using some appearance-related feature, in particular, symmetry has been extensively utilized. However, an in-depth analysis of the classification power of this feature is missing. As a first contribution of this paper, a thorough study of the classification performance of symmetry is presented within a Bayesian decision framework. This study reveals that the performance of symmetry-based classification is very limited. Therefore, as a second contribution, a new gradient-based descriptor is proposed for vehicle detection. This descriptor exploits the known rectangular structure of vehicle rears within a Histogram of Gradients (HOG)-based framework. Experiments show that the proposed descriptor outperforms largely symmetry as a feature for vehicle verification, achieving classification rates over 90%.

Comparison of 2.3 & 5 mega pixel (MP) resolution monitors when detecting mammography image blurring

Background - Image blurring in Full Field Digital Mammography (FFDM) is reported to be a problem within many UK breast screening units resulting in significant proportion of technical repeats/recalls. Our study investigates monitors of differing pixel resolution, and whether there is a difference in blurring detection between a 2.3 MP technical review monitor and a 5MP standard reporting monitor. Methods - Simulation software was created to induce different magnitudes of blur on 20 artifact free FFDM screening images. 120 blurred and non-blurred images were randomized and displayed on the 2.3 and 5MP monitors; they were reviewed by 28 trained observers. Monitors were calibrated to the DICOM Grayscale Standard Display Function. T-test was used to determine whether significant differences exist in blurring detection between the monitors. Results - The blurring detection rate on the 2.3MP monitor for 0.2, 0.4, 0.6, 0.8 and 1 mm blur was 46, 59, 66, 77and 78% respectively; and on the 5MP monitor 44, 70, 83 , 96 and 98%. All the non-motion images were identified correctly. A statistical difference (p <0.01) in the blurring detection rate between the two monitors was demonstrated. Conclusions - Given the results of this study and knowing that monitors as low as 1 MP are used in clinical practice, we speculate that technical recall/repeat rates because of blurring could be reduced if higher resolution monitors are used for technical review at the time of imaging. Further work is needed to determine monitor minimum specification for visual blurring detection.

Visual voice activity detection using frontal versus profile views

Visual activity detection of lip movements can be used to overcome the poor performance of voice activity detection based solely in the audio domain, particularly in noisy acoustic conditions. However, most of the research conducted in visual voice activity detection (VVAD) has neglected addressing variabilities in the visual domain such as viewpoint variation. In this paper we investigate the effectiveness of the visual information from the speaker’s frontal and profile views (i.e left and right side views) for the task of VVAD. As far as we are aware, our work constitutes the first real attempt to study this problem. We describe our visual front end approach and the Gaussian mixture model (GMM) based VVAD framework, and report the experimental results using the freely available CUAVE database. The experimental results show that VVAD is indeed possible from profile views and we give a quantitative comparison of VVAD based on frontal and profile views The results presented are useful in the development of multi-modal Human Machine Interaction (HMI) using a single camera, where the speaker’s face may not always be frontal.

Visual-Context Boosting for Eye Detection

Eye detection plays an important role in many practical applications. This paper presents a novel two-step scheme for eye detection. The first step models an eye by a newly defined visual-context pattern (VCP), and the second step applies semisupervised boosting for precise detection. VCP describes both the space and appearance relations between an eye region (region of eye) and a reference region (region of reference). The context feature of a VCP is extracted by using the integral image. Aiming to reduce the human labeling efforts, we apply semisupervised boosting, which integrates the context feature and the Haar-like features for precise eye detection. Experimental results on several standard face data sets demonstrate that the proposed approach is effective, robust, and efficient. We finally show that this approach is ready for practical applications.

Computational approaches to the visual validation of 3D virtual environments

Virtual environments can provide, through digital games and online social interfaces, extremely exciting forms of interactive entertainment. Because of their capability in displaying and manipulating information in natural and intuitive ways, such environments have found extensive applications in decision support, education and training in the health and science domains amongst others. Currently, the burden of validating both the interactive functionality and visual consistency of a virtual environment content is entirely carried out by developers and play-testers. While considerable research has been conducted in assisting the design of virtual world content and mechanics, to date, only limited contributions have been made regarding the automatic testing of the underpinning graphics software and hardware. The aim of this thesis is to determine whether the correctness of the images generated by a virtual environment can be quantitatively defined, and automatically measured, in order to facilitate the validation of the content. In an attempt to provide an environment-independent definition of visual consistency, a number of classification approaches were developed. First, a novel model-based object description was proposed in order to enable reasoning about the color and geometry change of virtual entities during a play-session. From such an analysis, two view-based connectionist approaches were developed to map from geometry and color spaces to a single, environment-independent, geometric transformation space; we used such a mapping to predict the correct visualization of the scene. Finally, an appearance-based aliasing detector was developed to show how incorrectness too, can be quantified for debugging purposes. Since computer games heavily rely on the use of highly complex and interactive virtual worlds, they provide an excellent test bed against which to develop, calibrate and validate our techniques. Experiments were conducted on a game engine and other virtual worlds prototypes to determine the applicability and effectiveness of our algorithms. The results show that quantifying visual correctness in virtual scenes is a feasible enterprise, and that effective automatic bug detection can be performed through the techniques we have developed. We expect these techniques to find application in large 3D games and virtual world studios that require a scalable solution to testing their virtual world software and digital content.